add includeList option

Author: jpinsonneau

docs/netobserv_cli.adoc

@@ -185,6 +185,7 @@ $ oc netobserv metrics [<option>]
 |--sports|                    filter on either of two source ports       | -
 |--tcp_flags|                 filter TCP flags                           | -
 |--interfaces|                interfaces to monitor                      | -
+|--include_list|              list of metric names to generate           | namespace_flows_total,node_ingress_bytes_total,node_egress_bytes_total,workload_ingress_bytes_total
 |===
 
 .Example running metrics capture for TCP drops

scripts/functions.sh

@@ -418,7 +418,7 @@ function copyFLPConfig() {
 
 # get network enrich stage
 function getNetworkEnrichStage() {
-  enrichIndex=$("$YQ_BIN" e -oj ".parameters[] | select(.name==\"enrich\") | document_index" "$json")
+  enrichIndex=$("$YQ_BIN" e -oj ".parameters[] | select(.name==\"enrich\") | path | .[-1]" "$json")
   enrichContent=$("$YQ_BIN" e -oj ".parameters[$enrichIndex]" "$json")
   enrichJson="${MANIFEST_OUTPUT_PATH}/enrich.json"
   echo "$enrichContent" >${enrichJson}
@@ -429,6 +429,19 @@ function overrideNetworkEnrichStage() {
   "$YQ_BIN" e -oj --inplace ".parameters[$enrichIndex] = $enrichJsonStr" "$json"
 }
 
+# get prometheus stage
+function getPromStage() {
+  promIndex=$("$YQ_BIN" e -oj ".parameters[] | select(.name==\"prometheus\") | path | .[-1]" "$json")
+  promContent=$("$YQ_BIN" e -oj ".parameters[$promIndex]" "$json")
+  promJson="${MANIFEST_OUTPUT_PATH}/prom.json"
+  echo "$promContent" >${promJson}
+}
+
+function overridePromStage() {
+  promJsonStr=$(cat "$promJson")
+  "$YQ_BIN" e -oj --inplace ".parameters[$promIndex] = $promJsonStr" "$json"
+}
+
 # update FLP Config
 function updateFLPConfig() {
   jsonContent=$(cat "$1")
@@ -631,6 +644,41 @@ function edit_manifest() {
     fi
     "$YQ_BIN" e --inplace ".spec.template.spec.nodeSelector.\"$key\" |= \"$val\"" "$manifest"
     ;;
+  "include_list")
+    # restrict metrics to matching items
+    copyFLPConfig "$manifest"
+    getPromStage
+
+    # list all matching metrics separated by new lines first
+    filteredMetrics=""
+    IFS=','
+    for match in $2; do
+      found=$("$YQ_BIN" -r ".encode.prom.metrics[] | select(.name | contains(\"$match\")).name" "$promJson")
+      if [ "${#filteredMetrics}" -gt 0 ]; then
+        filteredMetrics="${filteredMetrics}"$'\n'"${found}"
+      else 
+        filteredMetrics="$found"
+      fi
+    done
+
+    # then, format these for YQ filter function
+    echo "Matching metrics:"
+    match=""
+    IFS=$'\n'
+    for item in $filteredMetrics; do
+      echo " - $item"
+      if [ "${#match}" -gt 0 ]; then
+        match="$match,\"$item\""
+      else 
+        match="\"$item\""
+      fi
+    done
+
+    "$YQ_BIN" e --inplace ".encode.prom.metrics | filter(.name == ($match))" "$promJson"
+
+    overridePromStage
+    updateFLPConfig "$json" "$manifest"
+    ;;
   esac
 }
 
@@ -679,8 +727,12 @@ function check_args_and_apply() {
     *enable_pkt_drop) # Enable packet drop
       if [[ "$command" == "flows" || "$command" == "metrics" ]]; then
         defaultValue "true"
-        if [[ "$value" == "true" || "$value" == "false" ]]; then
+        if [[ "$value" == "true" ]]; then
           edit_manifest "pkt_drop_enable" "$value"
+          includeList="$includeList,workload_egress_bytes_total,namespace_drop_packets_total"
+        elif [[ "$value" == "false" ]]; then
+          # nothing to do there
+          echo
         else
           echo "invalid value for --enable_pkt_drop"
         fi
@@ -692,8 +744,12 @@ function check_args_and_apply() {
     *enable_dns) # Enable DNS
       if [[ "$command" == "flows" || "$command" == "metrics" ]]; then
         defaultValue "true"
-        if [[ "$value" == "true" || "$value" == "false" ]]; then
+        if [[ "$value" == "true" ]]; then
           edit_manifest "dns_enable" "$value"
+          includeList="$includeList,namespace_dns_latency_seconds"
+        elif [[ "$value" == "false" ]]; then
+          # nothing to do there
+          echo
         else
           echo "invalid value for --enable_dns"
         fi
@@ -705,8 +761,12 @@ function check_args_and_apply() {
     *enable_rtt) # Enable RTT
       if [[ "$command" == "flows" || "$command" == "metrics" ]]; then
         defaultValue "true"
-        if [[ "$value" == "true" || "$value" == "false" ]]; then
+        if [[ "$value" == "true" ]]; then
           edit_manifest "rtt_enable" "$value"
+          includeList="$includeList,namespace_rtt_seconds"
+        elif [[ "$value" == "false" ]]; then
+          # nothing to do there
+          echo
         else
           echo "invalid value for --enable_rtt"
         fi
@@ -718,8 +778,12 @@ function check_args_and_apply() {
     *enable_network_events) # Enable Network events monitoring
       if [[ "$command" == "flows" || "$command" == "metrics" ]]; then
         defaultValue "true"
-        if [[ "$value" == "true" || "$value" == "false" ]]; then
+        if [[ "$value" == "true" ]]; then
           edit_manifest "network_events_enable" "$value"
+          includeList="$includeList,namespace_network_policy_events_total"
+        elif [[ "$value" == "false" ]]; then
+          # nothing to do there
+          echo
         else
           echo "invalid value for --enable_network_events"
         fi
@@ -886,6 +950,14 @@ function check_args_and_apply() {
         echo "invalid value for --get-subnets"
       fi
       ;;
+    *include_list) # Restrict metrics capture
+      if [[ "$command" == "metrics" ]]; then
+        includeList="$value"
+      else
+        echo "--include_list is invalid option for $command"
+        exit 1
+      fi
+      ;;
     *) # Invalid option
       echo "Invalid option: $key" >&2
       exit 1
@@ -903,8 +975,10 @@ function check_args_and_apply() {
       echo
       exit 1
     fi
+  elif [[ "$command" = "metrics" ]]; then
+    # always restrict generated metrics
+    edit_manifest "include_list" "$includeList"
   fi
-
   yaml="$(cat "$manifest")"
   applyYAML "$yaml"
   if [[ "$outputYAML" == "false" ]]; then

scripts/generate-doc.sh

@@ -68,7 +68,7 @@ $ oc netobserv flows [<feature_option>] [<command_options>]
 features_usage
 collector_usage
 filters_usage
-specific_filters_usage
+flowsAndMetrics_filters_usage
 echo -e "|==="
 # flows example
 echo "
@@ -118,7 +118,8 @@ $ oc netobserv metrics [<option>]
 | Option | Description | Default"
 features_usage
 filters_usage
-specific_filters_usage
+flowsAndMetrics_filters_usage
+metrics_options
 echo -e "|==="
 # Metrics example
 echo "

scripts/help.sh

@@ -1,5 +1,8 @@
 #!/usr/bin/env bash
 
+# metrics includeList
+includeList="namespace_flows_total,node_ingress_bytes_total,node_egress_bytes_total,workload_ingress_bytes_total"
+
 # display main help
 function help {
   echo
@@ -107,11 +110,16 @@ function filters_usage {
   echo "  --tcp_flags:                  filter TCP flags                           (default: n/a)"
 }
 
-function specific_filters_usage {
-  # specific filters
+# specific filters for flows and metrics
+function flowsAndMetrics_filters_usage {
   echo "  --interfaces:                 interfaces to monitor                      (default: n/a)"
 }
 
+# specific filters for metrics
+function metrics_options {
+  echo "  --include_list:               list of metric names to generate           (default: $includeList)"
+}
+
 function flows_usage {
   echo
   echo "Netobserv allows you to capture flows from your cluster."
@@ -124,7 +132,7 @@ function flows_usage {
   echo
   echo "filters:"
   filters_usage
-  specific_filters_usage
+  flowsAndMetrics_filters_usage
   echo
   echo "options:"
   collector_usage
@@ -158,9 +166,12 @@ function metrics_usage {
   echo
   echo "filters:"
   filters_usage
-  specific_filters_usage
+  flowsAndMetrics_filters_usage
   echo "options:"
   script_usage
+  metrics_options
+  echo
+  echo "More information, with full list of available metrics: https://github.com/netobserv/network-observability-operator/blob/main/docs/Metrics.md"
 }
 
 function follow_usage {